This is the start of a blog on the work I'm doing with the Quantlib library. I find it useful to write stream of consciousness on what I'm doing. The point of my interest in Quantlib is to use this as a tool to do some research on convertible bonds in Chinese SOE's with the ultimate goal being to develop valuation models for Chinese state-owned enterprises. The first thing that I would like to do is to calculate implied credit risk for SOE convertible bonds.

Chinese corporations have some very interesting properties. The first is that they have an odd stock structure in which one third of the corporation consists of state-owned shares, one third of fa-gu (listed person shares) which are owned by legal persons, and one third of listed shares. Needless to say getting from the listed share price to the value of the corporation is a non-trivial exercise. This is made more difficult by the lack of transperancy in the system and the smallness of the stock market in which too much money is chasing far too few stocks.

Chinese corporate debt also has some interesting properties. Most work in the West with regard to convertible bonds focuses on the pricing of credit risk, since convertible bonds in the west tend to be used for startup growth corporations with the conversion provision as a sweetener. In a Chinese context, the bonds tend to be issued by large state owned enterprises, which are widely considered to have an implicit government guarantee against default.

The first part will include a lot of architectural cleanups of the Quantlib code to support convertible bond models.

-- JosephWang - 02 Feb 2005